Sandeep Midde

Technology & Engineering Transformation Leader

Leading technology transformation and engineering delivery for complex enterprise platforms.

Engineering executive with 20+ years of experience spearheading platform modernization and operational resilience. Most recently at Wells Fargo, directed critical engineering delivery for core finance platforms. Actively pioneering the next wave of enterprise tech through applied AI engineering, RAG evaluation, and agentic workflows.

20+ Yrs
Experience
$2.5T+
AUM supported
99.8%
Platform availability
Portrait of Sandeep Midde

Combining engineering depth with leadership accountability.

I started in software development and PeopleSoft consulting, advanced into technology architecture at Infosys, and spent 14 years at Wells Fargo, rising to Vice President. In that executive role, I directed platform strategy, governance, and delivery for core financial systems, collaborating closely with business and technology stakeholders on modernization, resilience, and compliance. Today, I bring that same engineering discipline to building GenAI applications hands-on.

VP
Wells Fargo
Platform strategy, governance and delivery
CTO
Sponsorship secured
For a build-versus-buy modernization
2×
Release cadence
Monthly to biweekly releases
30+
Distributed teams
Connected through one communication group

Delivering enterprise value.

Modernization, governance and operational resilience across two decades of enterprise delivery.

Finance Platform Modernization

Evaluated build-versus-buy options for replacing a legacy ERP platform and aligned Finance, business and technology stakeholders on the target state. Secured CTO sponsorship for the approach, which reduced projected total cost of ownership.

  • 70%Lower projected TCO
  • $1.2MReduced from $4M

Finance Workflow Transformation

Redesigned the computation and integration architecture of a critical Finance data-processing workflow, and architected a direct Bloomberg BEAP integration that retired a third-party intermediary.

  • 99%Lower data latency
  • 3h → 2mWorkflow time

Operational Resilience

Standardized service governance with SLAs and escalation frameworks, cutting MTTR by up to 60% and lifting availability from 99.2% to 99.8%. Scaled disaster recovery testing from annual to semi-annual, coordinated across 8 teams.

  • 99.8%Availability
  • <4 HrEPM recovery time

Regulatory Data Governance

Architected a COSO-aligned, audit-ready data retention framework that enforces Fed and bank-mandated compliance procedures by data classification, with automated file receipt validation and zero trust access controls.

  • COSOAligned framework
  • Zero TrustAccess controls

Delivery Operating Model

Restructured the engineering delivery model to instill clear module accountability and standardized governance. Optimized cross-team execution workflows to significantly improve time-to-market and build a more stable, resilient production environment.

  • 2×Release pace
  • 20%Fewer incidents

GenAI Adoption Evaluation

Directed a RAG feasibility study to shape the AI adoption roadmap, building an evaluation harness with no-answer detection. About 80% of responses were source-grounded, which informed the production go or no-go decision.

Increasing scope across two decades.

From PeopleSoft consulting and technology architecture to Vice President of Finance platforms, and now applied AI product development.

Jul 2026 to Present
Independent
AI R&D Lab

Principal AI Product Manager

Independently developed and deployed GenAI applications spanning conversational AI, knowledge management and agentic workflow automation, with hands-on ownership of product design, architecture and implementation.

AI Product StrategyProduct Lifecycle (0 to 1)Product Development & ValidationLLM Ops & Governance
Key highlights (4)
  • Built Krishna Uvaca, a RAG-based conversational application with grounded responses, multi-turn memory and vector retrieval. About 88% of answers were relevant and correctly cited across roughly 500 of its own test conversations, with no invented references found
  • Built Observer Prime, an agentic workflow and content automation platform with provider-agnostic LLM integration, resumable background jobs and 36 automated tests, currently in progress
  • Architected Elara, an AI knowledge management platform that uses AI only where understanding is needed and deterministic code for retrieval, keeping inference cost predictable, now live
  • Architected and deployed AutoPilot, a live beta multi-model job search assistant with predictive job matching and ATS analytics, backed by a 337-test suite
Oct 2012 to Jun 2026
Wells Fargo India
~14 years

Vice President

Directed platform strategy, engineering governance and delivery for core Finance FP&A platforms, aligning modernization initiatives with business objectives.

Platform GovernanceModernizationRisk & ComplianceTeam Leadership
Key highlights (9)
  • Directed a RAG feasibility study to shape the AI adoption roadmap, building an evaluation harness with no-answer detection and reaching about 80% source-grounded responses, which informed the production go or no-go decision
  • Drove a Finance platform modernization initiative, evaluating build-versus-buy options for legacy ERP replacement, aligning Finance, business and technology stakeholders, and securing CTO sponsorship, with projected TCO reduced by 70% ($4M to $1.2M)
  • Redesigned the delivery operating model into module-aligned teams with clear accountability and standardized governance, moving releases from monthly to biweekly and reducing deployment incidents by 20%
  • Strengthened operational excellence through standardized service governance, SLAs and escalation frameworks, reducing MTTR by up to 60% and improving availability from 99.2% to 99.8%
  • Scaled disaster recovery testing from annual to semi-annual, coordinating parallel server failovers across 8 teams and bringing EPM recovery time (RTO) under 4 hours
  • Architected a COSO-aligned, audit-ready data retention framework enforcing Fed and bank-mandated compliance procedures by data classification, with automated file receipt validation and zero trust access controls
  • Optimized IFRS-based General Ledger allocation processes supporting cost-center P&L reporting, reducing processing time by 58% (3.5 to 1.5 hours)
  • Transformed a critical Finance data-processing workflow integrating upstream data with PeopleSoft, redesigning computation and integration architecture to cut processing from 3 hours to under 2 minutes
  • Architected an extensible data infrastructure on the Bloomberg BEAP API to decommission a third-party intermediary, cutting data latency by 99% and establishing a blueprint for multi-system ingestion
Aug 2011 to Sep 2012
Oracle India
PeopleSoft consulting

Staff Consultant

PeopleSoft consulting on a global talent acquisition platform, from requirements to localization and performance.

RequirementsLocalization
Key highlights (2)
  • Drove requirements discovery with cross-functional stakeholders on the Talent Acquisition model, improving system performance by about 20% and stabilizing workflows
  • Delivered multilingual support for Inspira, the UN talent acquisition platform, enabling regional configurability and compliance for rollout across 190+ countries
Jul 2006 to Jul 2011
Infosys
Charles Schwab · Dow Jones

Technology Architect

PeopleSoft Finance delivery for Charles Schwab and Dow Jones under CMMI Level 5 governance.

ERP DeliveryCMMI Level 5Client Leadership
Key highlights (5)
  • Managed cross-functional delivery and the end-to-end SDLC for the Charles Schwab account across design, upgrade and support engagements
  • Led Finance transformation delivery across P2P, O2C and R2R (GL, AP, AR, Asset Management, Purchasing), coordinating the upgrade of 900+ custom objects through implementation, deployment and hypercare
  • Owned delivery governance for recurring production incidents under CMMI Level 5, cutting the ticket backlog by up to 40% during financial close
  • Revived a stalled Finance transformation program at Dow Jones by analyzing undocumented legacy workflows and resolving delivery dependencies, unlocking the next phase of project funding
  • Chaired the LOB Communication Group, uniting 30+ distributed teams through structured knowledge-sharing

Leadership & Technology Principles.

Value and Cost Before Commitment

Technology decisions start with business value and total cost of ownership. Cost models and technical feasibility are validated before capital is committed, and AI adoption is judged the same way, on evidence, controls and measurable value.

Ownership and Predictable Delivery

Reliability comes from clear team ownership, defined service levels and rehearsed recovery, not from fast reactive troubleshooting. Engineering quality and release speed are balanced through standard governance.

Applying enterprise engineering principles to practical GenAI solutions.

Independent product development in retrieval, agentic and knowledge systems, from architecture to working build. These are personal projects, not enterprise deployments, and each one is written up in full.

ShippedKṛṣṇa Uvāca cover

GenAI knowledge product

Kṛṣṇa Uvāca: Grounded Cosmic Counsel

A conversational interface enabling direct dialogue with the persona of Lord Kṛṣṇa. Grounded in authenticated Bhagavad Gītā scriptures, it pairs precise textual citations with philosophical interpretations.

  • RAG
  • Vector retrieval (ChromaDB)
  • Source citations

~88% accuracy across ~500 of its own test conversations

Strict out-of-scope rejection

In ProgressObserver Prime cover

Agentic orchestration product

Observer Prime: Unified Agentic Orchestration

A multi-agent orchestration product that automates complex, multi-step workflows, with human checkpoints that preserve strategic judgment.

  • Agentic workflows
  • Provider-agnostic LLM
  • Human-in-the-loop

Central governance across every workflow

Validated end to end on a zero-cost mock

ShippedProject Elara cover

Intelligent knowledge workspace

Project Elara: Contextual Knowledge OS

A personal workspace that turns PDFs, articles and digital clutter into a structured, searchable library.

  • AI knowledge management
  • Automated ingestion
  • Deterministic retrieval

Low-overhead data architecture

Validated with live sign-in and AI calls

Live BetaAutoPilot cover

AI job-search assistant

AutoPilot: An AI Assistant for the Job Hunt

A local-first assistant that finds, scores and prepares job applications, while every submit and send stays with me.

  • Multi-model LLM
  • Predictive matching
  • Human-in-the-loop

A 4 to 5 hour manual search, now a short list in minutes

Every submit and send stays a human click

All Build Logs

Three areas of depth.

Technology Leadership & Transformation

Engineering Leadership & People Management Technology Strategy & Roadmaps Enterprise Architecture (TOGAF) Platform Modernization Program & Delivery Governance Executive Stakeholder Management Risk, Compliance & Audit Readiness (COSO · CMMI L5)

Hands-on capabilities

Engineering & Platform Delivery

Finance & ERP Platforms (PeopleSoft · Oracle) CI/CD & Release Management Cloud Platforms (Azure · AWS) Python API & Enterprise Integration (REST · Apigee) Production Operations & Resilience SOA & Event-Driven Architecture (Kafka)

AI, GenAI & Intelligent Automation

Generative AI & LLM Applications RAG & Vector Retrieval (ChromaDB) Agentic AI & Workflow Automation AI Evaluation & Guardrails Prompt Engineering & Responsible AI Multi-Model LLM Orchestration AI Frameworks (LangChain · LangGraph)

Certifications and education.

Certifications

  • Project Management Professional (PMP) PMI
  • Oracle Certified PeopleSoft Technical Consultant Oracle
  • Azure Fundamentals (AZ-900) Microsoft
  • Enterprise Product Management Fundamentals Microsoft
  • AI Security & Governance Securiti
  • Agentblazer Champion 2026 Salesforce

Education

  • Post Graduate, AI & Machine Learning IIIT Bangalore
  • MBA, IT & Financial Accounting ICFAI
  • B. Pharmacy (Hons.) BITS Pilani

Let's connect.

Open to conversations on technology leadership, engineering transformation, enterprise platforms and applied AI.